Interaction of Partially Structured States of Acidic Fibroblast Growth

Interaction of Partially Structured States of Acidic Fibroblast Growth Factor with Phospholipid Membranes ... Click to increase image size Free first ...
0 downloads 0 Views 980KB Size
Biochemistry 1995,34, 9913-9920

9913

Interaction of Partially Structured States of Acidic Fibroblast Growth Factor with Phospholipid Membranes Henryk Mach and C. Russell Middaugh* Department of Pharmaceutical Research, WP78-302, Merck Research Laboratories, West Point, Pennsylvania 19486 Received February 13, 1995; Revised Manuscript Received May 4 , 1995@

ABSTRACT: Although acidic fibroblast growth factor (aFGF) lacks a conventional signal sequence, it is

often found complexed to sulfated proteoglycans on the external surface of cells. The protein also forms a “molten globule”-like state at neutral pH and physiological temperatures as well as at acidic pH in the presence of physiological ionic strength or moderate quantities of polyanions. These states display a marked tendency to aggregate. Such observations suggest that related partially structured states might be involved in the membrane translocation of aFGF. To explore this hypothesis, we examined the interaction of this growth factor with lipid vesicles as well as the effect of such surfaces on the structure of the protein. We find that these states interact with negatively charged but not neutral phosholipid unilammelar vesicles at acidic pH, inducing bilayer disruption. The rate of leakage of a liposome-entrapped fluorescent probe is proportional to the logarithm of the aFGF concentration, suggesting competition between protein self-association and membrane binding. Liposome leakage can be also induced at neutral pH by partial unfolding of aFGF at or above physiological temperature in contrast to most control proteins. The importance of partially folded hydrophobic surfaces in aFGF self-association and membrane binding is further suggested by the fact that thermally unfolded aFGF does not aggregate, in contrast to states observed at intermediate temperatures or transiently during unfolding at high temperatures. In contrast to heparin, a polyanion which stabilizes the native structure of aFGF, negatively charged phospholipid membranes appear to enhance the disruption of aFGF tertiary structure at submicellar concentrations of sodium dodecyl sulfate but stabilize the remaining secondary structure. Thus negatively charged lipid bilayers appear to interact with partially structured states of aFGF by preferential binding of both its apolar and charged surfaces to complementary regions of the lipid bilayer. Such interactions may play a role in the membrane translocation of this growth factor.

Acidic fibroblast growth factor (aFGF;’ FGF-1) is a 16kDa protein possessing mitogenic, angiogenic, and chemotactic stimulatory activities [reviewed by Burgess and Maciag (1989)l. The three-dimensional structure of the bovine form of the protein has been determined by X-ray crystallography and the human form by NMR (Pineda-Lucena et al., 1994) and consists of 12 antiparallel P-strands with pseudo 3-fold symmetry (Zhu et al., 1991). In the absence of stabilizing polyanions, aFGF is significantly unfolded at physiological temperatures (>30 “C) (Copeland et al., 1991; Dabora et al., 1991) and requires polyanions such as heparin to maintain its native structure and biological activity (Gospodarowicz & Cheng, 1986; Volkin et al., 1993). At low pH, the presence of small amounts of polyanionic ligands or moderate amounts of salts induces formation of partially folded states in aFGF, characterized by high secondary but low tertiary structure content. At neutral pH, intermediate amounts of chaotropic agents or slightly elevated temperature (’30 “C) impose similar “molten globule”-like conformational states which display noncooperative unfolding transitions (Mach et al., 1993). Unlike the native and fully unfolded states, these partially structured conformations exhibit very low solubility, resulting in irreversible aggrega~~~~~~~~~

~~

~~

* To whom correspondence should be addressed. Phone: (215) 652-

3438. Fax: (215) 652-5299. Abstract published in Advance ACS Abstrucrs, July 15, 1995. @

Abbreviations: aFGF, acidic fibroblast growth factor; ANS, 8-aniline-1-naphthalenesulfonate;DOPG, dioleoylphosphatidylglycerol; DOPC, dioleoylphosphatidylcholine. 0006-2960/95/0434-99 13$09.00/0

tion of aFGF. This tendency to aggregate as well as the ability of the protein to bind amphiphilic dyes (e.g., ANS) suggest that partially folded states of aFGF may have the potential to interact with lipid bilayers. “Molten globule” states have been implicated in membrane interactions of several proteins [e.g., Bychkova et al. (1988) and van der Goot et al. (1991)l. The membrane insertion and translocation of the majority of proteins appears to be mediated by a hydrophobic signal peptide which is subsequently enzymatically cleaved to yield the mature functional protein. Signal peptides also generally possess several positive charges proximate to an apolar region arranged in such a manner that positively charged and apolar faces are created by a-helical secondary structure [for reviews, see Kaiser and Kezdy (1987), Attardi and Schatz (1988), Wanner and Neupert (1990), and Glick and Schatz (1991)l. Although aFGF lacks a conventional signal sequence, it has been shown to translocate through biological membranes when fused with the A-fragment of diptheria toxin (Wiedlocha et al., 1992). The requirement of low pH and blockage of translocation by stabilizing polyanions suggests the possible involvement of “molten globule” states of aFGF in this process. Furthermore, the ability of the aFGF-toxin conjugate to stimulate DNA synthesis in the absence of functional aFGF membrane receptors is consistent with such a mode of transport (Wiedlocha et al., 1994). In the present study, we characterize the interaction of aFGF with phos0 1995 American Chemical Society

Mach and Middaugh

9914 Biochemistry, Vol. 34, No. 31,1995 pholipid vesicles in an exploration of this hypothesis. We find that aFGF-induced disruption of membrane structural integrity coincides with the formation of partially structured states at physiological temperature (> 30 "C) possessing the characteristics of "molten globules". These states are formed at low pH ( 30 "C) or at low pH in the presence of anions including certain phospholipids, the

9918 Biochemistry, Vol. 34, No. 31, 1995 protein manifests little evidence of stable tertiary structure although its secondary structure remains essentially intact. These states are significantly apolar as evidenced by the binding of hydrophobic dyes and display a marked tendency to aggregate. In this work, we show that such states are also able to interact with lipid bilayers as evidenced by their ability to induce the release of encapsulated dyes from negatively charged liposomes as well as cause aggregation and fusion of the same vesicles. Several experiments hint at the nature of the interaction of aFGF with these lipid bilayers. The leakage of the fluorescent probe carboxyfluorescein at a threshold near pH 4.5 parallels the cell membrane translocation of an aFGFdiphtheria toxin conjugate. A dramatic increase in the normally quenched tryptohan fluorescence in this same pH range suggested significant disruption of aFGF native structure during translocation of this conjugate (Wiedlocha et al., 1992). In contrast, aFGF in its native, polyanionstabilized compact state appears to be unable to penetrate lipid bilayers. It seems probable, however, that in the absence of high affinity polyanions and at low ionic strength and low temperatures which preserve native structure, aFGF can bind to liposomes by the virtue of their high negative charge density. It has previously been shown that aFGF can bind to wide variety of polyanions and that charge density appears to be the major determinant of affinity (Volkin et al., 1993). Nevertheless, the growth factor must associate more tightly with cell surface proteoglycans, such as heparan sulfate than with phospholipids, since the former are found to be the primary storage depot of the protein on cell surfaces (Vlodavsky et al., 1991). Several events occur upon lowering the pH of a protein solution near or below the protonation point of aspartic and glutamic acid side chains. The disappearance of compensating negative charges increases the repulsion between positively charged side chains, such as lysine and arginine. This effect is especially pronounced when the ionic strength is lowered. Under these conditions, aFGF is very soluble and manifests a CD spectrum characteristic of a randomly coiled polypeptide (Mach et al., 1993). Addition of counterions, however, neutralizes long-range charge repulsion between side chains and allows the formation of a partially folded state. At neutral pH, ambient temperature, and low ionic strength aFGF is significantly unfolded (not illustrated), presumably due to the same repulsive forces. In addition, the disappearance of negative charge increases the net positive charge and, consequently, the affinity for negatively charged phospholipid surfaces (Kim et al., 1991). Upon binding, the well defined positively charged patches on the surface of aFGF presumably directly interact with the lipid bilayer surface while some of the more apolar regions of the protein may insert at least partially into the bilayer. By analogy, both positively charged and apolar residues of phospholipase A2 appear to be involved in this proteins interaction with membranes (Scott et al., 1990). Aggregation of such partially embedded protein molecules within the lipid bilayer may then result in membrane disruption and the leakage of any encapsulated contents. In micelles and lipid bilayers composed of either neutral or charged amphiphiles, even if the bulk pH changes by a large amount, dissociable groups on the surface are buffered by the surrounding lipid heads (Scarlata & Rosenberg, 1990). After initial ionization, other phosphate protons in bilayers

Mach and Middaugh can stabilize the negatively charged surface through an extensive hydrogen-bonding network in which the protons are delocalized over two phosphate groups (Eibl, 1983). Consequently, stabilization of the protonated state of groups by the negatively charged surface occurs. For example, insertion of potentially negatively charged 6-decanoylnapthol or 3-hexadecanoylcoumarin into SDS micelles shifts the probes dissociation constant upward by 2 pH units (Scarlata & Rosenberg, 1990). This suggests that, as a protein approaches a negatively charged bilayer surface, the acidic groups of the protein may lose their charge promoting protein unfolding and membrane insertion. At elevated temperatures, where the formation of the “molten globule” state can be achieved without low pH, naturally occurring basic patches (e.g., residues 100-128) of aFGF (Zhu et al., 1993; Chavan et al., 1994; Pineda-Lucena et al., 1994) together with the apolar and protonation inducing properties of the bilayer may induce insertion. The importance of negative charge density appears to be more crucial at neutral than at low pH (Figures 1 and 8). This is presumably due to the presence of negatively charged groups remaining on aFGF upon membrane binding under these conditions. Unlike many of the channel-forming membrane proteins, which are able to functionally insert in a monomeric or limited oligomeric state, aFGF’s interaction with membranes appears to be in competition with self-association, as judged from the logarithmic dependence of liposome leakage rates on aFGF concentration. One of the best studied watersoluble proteins capable of membrane insertion and ion channel formation, colicin A, manifests single-hit insertion kinetics (Me1 & Stroud, 1993). Its three-dimensional structure in aqueous solution has been determined (Parker et al., 1989, 1992) and is composed of a bundle of 10 helices arranged in three layers. The kinetics of insertion of colicin A into lipid bilayers at low pH also correlates with the appearance of a “molten globule” state (van der Goot et al., 1991). After insertion, the protein apparently remains in this conformation (Muga et al., 1993). There is good evidence that most of the protein resides at the surface of the bilayer with only helices 8 and 9 inserted into the bilayer’s apolar interior (Massotte et al., 1993). The area at the bilayer surface occupied by the bound colicin A appears to be larger than the cross-sectional area of the native compact solution form as a consequence of this conformational change. A similar result can be deduced from the average lipid bilayer area occupied by one aFGF molecule under saturating conditions, estimated to be about 21 nm2, or a square of dimension 4.6 nm. Native aFGF has a hydrodynamic diameter of approximately 3.7 nm (equivalent to a crosssectional area of about 11 nm2),assuming spherical geometry, implying the possibility of significant expansion of the bound growth factor, although these estimates are sufficiently crude that no definitive conclusion should be drawn. There are major changes in the fluorescence emission of the protein’s single tryptophan residue upon interaction with liposomes implying transfer of at least this residue to a more apolar environment. Fluorescence quenching experiments also detect a decrease in solvent accessibility of this side chain, although binding of heparin to aFGF in solution causes an even greater loss in accessibility (but no change in overall intrinsic fluorescence properties). The precise nature of this conformational change is difficult to ascertain at this point. One possibility is that the planar lipid bilayer may force the

Interaction of aFGF with Membranes protein into a somewhat more extended conformation, resulting in greater tryptophan exposure. The lack of energy transfer between the tryptophan residue and pyrenyl dodecanoic acid embedded in the bilayer interior is consistent with some Trp solvent exposure of the liposome-bound aFGF. Use of the protein’s single indole group as a probe of overall protein structure, however, could be misleading in this regard. Unlike numerous natural and synthetic peptides that form amphiphilic a-helices in the presence of lipid membranes, aFGF bound to liposomes appears to maintain its predominantly /3-strand structure (Mach et al., 1993). Although it is not possible to perform definitive CD measurements under exactly the same conditions as the dye release experiments, the observed distorted spectra as well as more interpretable FTIR studies argue that substantial /3-type secondary structure remains in the liposome-associated protein. These results are supported by more unambiguous CD and fluorescence studies in SDS which clearly show that the anionic lipid bound protein possesses extensive secondary structure but dramatically decreased tertiary contacts (as evidenced by the loss of the 228 nm CD band and large increases in the 350/300 nm fluorescence ratio). Another possibility is that a segment of the polypeptide chain of aFGF become extruded from the major body of the protein and serves as a membrane interaction vehicle like the situation observed with the influenza hemagluttinin (Bullough et al., 1994). Many of the proteins and peptides known to insert into lipid bilayers require acidic conditions for their functional activity (e.g., Sandvig & Olsnes, 1988; Kono et al., 1990; Levy-Mintz, & Kielian, 1991; Stegmann et al., 1991; Sanyal et al., 1993; Collins & Cha, 1994). Similarly, a disruption of native structure and formation of a “molten globular” state facilitated by low pH was previously shown to be a prerequisite for membrane insertion and translocation of aFGF conjugated to the A-fragment of diptheria toxin (Wiedlocha et al., 1992). In this study, we demonstrate that at neutral pH thermally induced “molten-globular” states of aFGF are also capable of functional membrane insertion. This may represent a rather general phenomenon. For example, conformational changes of influenza hemagglutinin related to fusion can also be triggered by high temperature at neutral pH (Ruigrok et al., 1986). In these cases, hydrogen ions are thought to act only to enable the conformational changes and not to directly participate in membrane insertion events per se. The requirement for the presence of negatively charged lipids for aFGF to insert into the bilayer at neutral pH appears to be more important than at lower pH, presumably as a consequence of a less favorable charge balance. Of some interest is the question of whether there exists a correlation between the aggregation tendencies of partially folded states and membrane insertion propensity. A simplistic consideration of the forces involved in protein folding would suggest that minimization of unfavorable entropic effects by shielding the apolar regions from water might indeed be a common feature involved in the formation of “molten globules” and lipid bilayer insertion. In particular, in this work we show that thermally unfolded aFGF does not aggregate. To demonstrate this, the concentrated protein unfolded by a chaotropic agent was introduced into bulk solutions at various temperatures. Only over a relatively narrow range of temperatures (45 to 55 “C) could aggregation

Biochemistry, Vol. 34, No. 31,1995 9919 be observed. This is exactly the temperature range over which the protein’s fluorescence properties indicate disruption of tertiary structure but far-UV CD spectra imply that a significant amount of secondary structure still persists (Mach et al., 1993). These, of course, are the hallmarks of “molten globule” states, which have previously been linked to heatinduced aggregation (Chrunyk et al., 1993; Fischer et al., 1993). The temperature-dependent aggregation profile observed in this work reinforces the idea that intermediate rather than high (Le., unfolding) temperatures are responsible for most “heat denaturation” of proteins. Acidic FGF and perhaps most water soluble proteins are probably very soluble at temperatures high enough to induce more extensively unfolded states. In a limited predenaturing temperature region, however, partially unfolded (tertiary structuredisrupted) states appear, and the apolar surfaces of these molten-globule states induce aggregation. Since the size of the aggregates frequently reach the submicron level, the unfavorable surface-to-volume ratio and possibly polypeptide chain entanglement cause the apparent irreversibility that is usually observed. In conclusion, these results suggest the following simple model: the strong electrostatic attraction between the negatively charged phospholipid vesicles and positively charged regions of aFGF are responsible for the formation of an initial complex between the two entities. Unlike the interaction with heparin and other related polyanions, the binding of aFGF to liposomes does not seem to induce the formation of native structure. Rather, the liposomes appear to stabilize intermediate “molten globule” states of aFGF presumably through apolar interactions between protein and lipid side chains. This does not, however, seem to involve extensive penetration of the majority of the protein into the lipid bilayer as evidenced by energy transfer results. Most simply, one can envision penetration of perhaps one or a few elements of the protein’s secondary structure into the bilayer in much the same way that a number of fusion and pore-forming proteins mediate their physiological activity (e.g., Doms et al., 1985; Novick & Hoekstra, 1988; Stegmann et al., 1991; Bullough et al., 1994; Yu et al., 1994). Thus, the lipid bilayers can be viewed as simply stabilizing one or a few substates of the rather dynamic MG states. It is not clear how this initial interaction might lead to actual translocation. This process may be facilitated in vivo by other (perhaps chaperone-like) proteins although it is possible that the phenomenon described in this work can itself lead to translocation competent species in appropriate environments. It must be emphasized, however, that a role for such states in the transport of this protein remain hypothetical.

REFERENCES Attardi, G . , & Schatz, G. (1988) Annu. Rev. Cell Biol. 4 , 289333. Bruggennann, E. P., & Kayalar, C. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 4273-4276. Bullough, P. A., Hughson, F. M., Skehal, J. J., & Wiley, D. C. (1994) Nature 371, 37-43. Burgess, W. H., & Maciag , T. (1989) Annu. Rev. Biochem. 58, 575-606. Bychkova, V. E., Pain, R. H., & Ptitsyn, 0. B. (1988) FEBS Lett. 238, 23 1-234. Chavan, A. J., Haley, B. E., Volkin, D. B., Marfia, K. E., Verticelli, A. M., Bruner, M. W., Draper, J. P., Burke, C. J., & Middaugh, C. R. (1994) Biochemistiy 33, 7193-7202.

9920 Biochemistry, Vol. 34, No. 3 1 , 1 9 9 5 Chrunyk, B. A., Evans, J., Lillquist, J., Young, P., & Wetzel, R. (1993) J. Biol. Chem. 268, 18053-18061. Collins, D., & Cha, Y. (1994) Biochemistry 33, 4521-4526. Copeland, R. A,, Ji, H., Halfpenny, A. J., Williams, R. W., Thompson, K. C., Herber, W. K., Thomas, K. A,, Bruner, M., Ryan, J. A., Marquis-Omer, D., Sanyal, G., Sitrin, R. D., Yamazaki, S., & Middaugh, C. R. (1991) Arch. Biochem. Biophys. 289, 53-61. Dabora, J. M., Sanyal, G., & Middaugh, C. R. (1991) J.Biol.Chem. 266, 23637-23640. Doms, R. W., Helenius, A., &White, J. (1985) J . Biol.Chem. 260, 2973-2980. Eibl, H. (1983) Membrane Fluidity in Biology (Alaoia, Ed.) Vol. 2, pp 217-236, Academic Press, New York. Fisher, B., Sumner, I., & Goodenough, P. (1993) Arch. Biochem. Biophys. 306, 183-187. Freire, E., Markello, T., Rigell, C., & Holloway, P. W. (1983) Biochemistry 22, 1675-1680. Glick, B. S., & Schatz, G. (1991) Annu. Rev. Genet. 25, 22-44. Gospodarowicz, D., & Cheng, J. (1986) J . Cell. Physiol. 128,475484. Kaiser, E. T., & Kezdy, F. J. (1987) Annu. Rev. Biophys. Biophys. Chem. 16. 561-581. Kim, J., Mosior, M., Chung, L., & McLaughin, S, (1991) Biophys. J . 60, 135-148. Kono, K., Kimura, S., & Imanishi, Y. (1990) Biochemistry 29, 363 1-3637. Koppel, D. E. (1972) J . Chem. Phys. 57, 4814-4820. Levinthal, F., Todd, A. P., Hubbel, W. L., & Levinthal, C. (1991) Proteins: Struct., Funct., Genet. 11, 254-262. Levy-Mintz., P., & Kielian, M. (1991) J . Virol. 65, 4292-4300. Mach, H., & Middaugh, C. R. (1994) Arch. Biochem. Biophys. 309, 36-42. Mach, H., Ryan, J. A., Burke, C. J., Volkin, D. B., & Middaugh, C. R. (1993) Biochemistry 32, 7703-7711. Massotte, D., Yamamoto, M., Scianimanico, S., Sorokine, O., van Dorsselaer, A., Nakatani, Y., Ourisson, G., & Pattus, F. (1993) Biochemistry 32, 15787-15794. Mel, S. F., & Stroud, R. M. (1993) Biochemistry 32, 2082-2089. Michelangeli, F., East, J. M., & Lee, A. G. (1990) Biochim. Biophys. Acta 1025, 99-108. Muga, A., Gonzales-Manas, J. M., Lakey, J. H., Pattus, F., & Surewicz, W. K. (1993) J . Biol. Chem. 268, 1553-1557. Novick, S. L., & Hoekstra, D. (1988) Proc. Natl. Acad. Sci. U.S.A. 85, 7433-7437.

Mach and Middaugh Parker, M. W., Pattus, F., Tucker, A. D., & Tsernoglou, D. (1989) Nature 337, 93-96. Parker, M. W., Tucker, A. D., Tsemoglou, D., & Pattus, F. (1990) Trends Biochem. Sci. 15, 126-129. Pfanner, N., & Neupert, W. (1990) Annu. Rev. Biochem. 59,331353. Pineda-Lucena, A., Jimenez, M. A,, Nieto, J. L., Santoro, J., Rico, M., & Gimenez-Gallego, G. (1994) J . Mol. Biol. 242, 81-98. Ruigrok, R. W., Martin, S. R., Wharton, S. A., Skehel, J. J., Bayley, P. M., & Wiley, D. C. (1986) Virology 155, 484-497. Sandvig, K., & Olsnes, S. (1988) J.Biol. Chem. 263, 12352-12359. Sanyal, G., Marquis-Omer, D., O’Brien Gress, J., & Middaugh, C. R. (1993) Biochemistry 32, 3488-3497. Scarlata, S. F., & Rosenberg, M. (1990) Biochemistry 29, 1023310240. Scott, D. L., White, S. P., Otwinowski, Z., Yuan, W., Gelb, M. H., & Sigler, P. B. (1990) Science 250, 1541-1546. Small, D. M. (1986) in Handbook ofLipid Research: The Physical Chemistry of Lipids: From Alkanes to Phospholipids (Hanahan, D. J., Ed.) Plenum Press, New York. Stegmann, T., Delfino, J. M., Richards, F. M., & Helenius, A. (1991) J . Biol. Chem. 266, 18404-18410. van der Goot, F. G., Gonzales-Manas, J. M., Lakey, J. H., & Pattus, F. (1991) Nature 354, 408-410. Vlodavsky, I., Fuks,. Z., Ishai-Michaeli, R., Bashkin, P., Levi, E., Korner, G., Bar-Shavit, R., & Klagsbrun, M. (1991) J . Cell. Biochem. 45, 167-176. Volkin, D. B., Tsai, P.-K., Dabora, J. M., Gress, 0. J., Burke, C. J., Linhardt, R. J., & Middaugh, C. R. (1993) Arch. Biochem. Biophys. 300, 30-41. Wendt, L. (1970) J. Bacteriol. 104, 1236-1241. Wiedlocha, A., Madshus, I. H., Mach, H., Middaugh, C. R., & Olsnes, S. (1992) EMBO J . 11, 4835-4842. Wiedlocha, A., Falnes, P. D., Madshus, I. H., Sandvig, K., & Olsnes, S. (1994) Cell 76, 1039-1051. Yu, Y. G., King, D. S., & Shin, Y.-K. (1994) Science 266, 274276. Zhu, X., Komiya, H., Chirino, A., Faham, S., Fox., G. M., Arakawa, T., Hsu, B. T., & Rees, D. C. (1991) Science 251, 90-93. Zhu, X., Hsu, B. T., & Rees, D. C. (1993) Structure 1 , 27-34. BI950323L